Independent local planning aid · source checked September 2, 2026

Is your Enterprise Frontier Safeguards plan operationally ready?

Review the controls that sit behind the privacy claim: exact eligibility, phased rollout, transition ZDR, customer-owned storage, keys, automated monitoring, customer reviewers, cloud cost, failure tests, and rollback.

Runs locallyNo provider connectionNot compliance advice

Readiness evidence

Check an item only when a named owner can show current evidence.

Hard blockers

Planning result

Complete the evidence review

Select only controls backed by the live account, architecture, policy, and test record.

Transition: eligible Fable 5 and Fable 5.1 customers can use ZDR until EFS is ready.

Target: customer-controlled cloud storage plus opt-in keys and automated review.

Human review: flags go to the customer; Anthropic human review is not required.

Cost: no Anthropic EFS fee; cloud operations and egress can still be billed.

Interpret the result

A blocker means the plan relies on an unsupported availability, coverage, or privacy inference. Resolve it before moving sensitive traffic or depending on a future control.

A candidate result means all eight planning gates have evidence and no blocker is asserted. It does not prove that Anthropic has enabled EFS or that regulators, legal, security, and operations have approved production use.

Customer-owned storage is an operating system, not a checkbox. Test key loss, access denial, late logs, deletion, cross-account access, backup recovery, and incident evidence before launch.

Automated monitoring still needs human accountability. Define who can see a flag, what they may inspect, how false positives are cleared, when access is revoked, and how incidents are reported.

Supported-surface evidence

Anthropic says EFS will support Claude Code, Claude Enterprise, the Claude Platform, Amazon Bedrock, Claude Platform on AWS, Google's Agent Platform, and Microsoft Foundry. A named future surface is not proof that one account, model, region, or contract is enabled today.

Fable 5.1 record

API ID, public price, cache reads, safeguards, clouds, and migration.

Mythos 5.1 boundary

Trusted programs and why Fable transition language must not be inferred.

Safeguard decision

Compare 5 and 5.1 lifecycle, access, redirects, price, and data controls.

Enterprise Frontier Safeguards questions

It checks whether a team has documented eligibility, rollout, transition ZDR, customer storage, encryption keys, automated monitoring, customer review, audit, cost, tests, and rollback. It runs locally and does not inspect or change a provider or cloud account.
Anthropic says EFS can store activity data used for monitoring in cloud infrastructure controlled by the customer. The customer can govern encryption keys, access policies, and audit logs, while automated monitoring can analyze a rolling traffic window.
Anthropic says flags go directly to the customer and no Anthropic human review is required. The customer still needs authorized reviewers, a triage procedure, evidence handling, incident response, and false-positive resolution.
Anthropic says it does not charge for EFS and that EFS does not change model behavior, API pricing, or rate limits. The customer's cloud provider can still bill storage, reads, writes, and data egress, and the organization bears its own security and review costs.
No. It means the planning evidence is represented. Live eligibility, rollout, contract, technical configuration, legal and security review, test results, regulatory fit, and production approval still require accountable evidence.

Official references: Enterprise Frontier Safeguards announcement and Fable 5.1 and Mythos 5.1 launch. Recheck live rollout and account evidence before use.